the analysis in scott's most recent post i think is completely wrong. should i write a post explaining why and what i think is going on instead
Conversation
That it's possible for fear to be non-aversive doesn't seem to affect his central point, I think? Even if it's possible for it to be non-aversive (and I agree it is), for nearly all people there are contexts in which it isn't, and that's enough to drive motivated reasoning.
2
4
the point i’m arguing against is scott’s claim that there are two kinds of learning and that reinforcement learning could be “mis-applied” - there’s just one kind of learning and it gets applied to internal states sometimes and that’s enough to drive motivated reasoning
1
1
Okay. I don't think I follow you. :)
1
1
i’m so confused but tbc this is the sentence i think is the most wrong. maybe i should just write the post 😅
1
2
Hmm okay, I would likely word things a little differently from Scott, but the claim that the brain has one type of learning that's like RL and another that's more like prediction seems reasonable to me, even if they definitely get somewhat intertwined too
2
Replying to
i think there’s one type of learning that isn’t quite like either (PCT is the model i currently like the best but i don’t understand well enough to use in detail) but push comes to shove is a bit more like RL. but again, maybe i should just write the post…
Replying to
If it helps I can keep responding to you with questions/comments so you keep going "oh I really should write that post" until it turns into "oh ffs *writes post* "